Lights... Camera... Action!

THE 3-D PROCESS

When a client asks us to build an animation, our first question is "who is the audience?" The second question is "what do you want explained?" And the third is "why?" It's our loose interpretation of the Konstantin Stanislavski method of acting: who am I, what do I want, and where am I going? In order to simplify, one must first understand in depth. The surest way to really find out what's current in any given field is to start a debate. We've been very fortunate in finding bench scientists with a knack for relaying concepts in their field to the public. That's the first filter. From there, we become the lens, focusing that information down until we have the core of what we're trying to convey, and no more. These animations are never meant to be the definitive tome on any given subject, but to stimulate curiosity to find out more.

Take the example of a US biotech company who in 1998 needed to have an engaging descriptive piece about proteomics that spoke to a few audiences – lay investors and scientists. At the time we were asked to do the animation, the field of proteomics was coming into its own. The task we were handed was to develop a storyline that quickly and effectively gave the viewer a synopsis of what proteins are, and the myriad roles they play in bodily function.

A major question during this phase was, could we depict the interactions between the "characters" in an analogous manner, or should it be more representational – using less-detailed objects, and focus more on the interaction between the objects to tell the story? Together with company scientists, we developed the idea of using real-world analogies to describe protein function: forklifts and dump trucks for transporter proteins, radar dishes for signaling proteins, and power plants for mitochondrial proteins, for example. This went through a series of refinements until we had similes which best resonated.

Overflowing with information and ideas, we began writing the storyline (a brief one- to two-page synopsis) and script. The script, necessary even if no voiceover is added, is usually a collaborative effort between Arkitek and the client; they know their subject matter best, while we lend our thoughts on how it will play best to the target audience.

Photo: Doug Huff

At that point we were ready to create the storyboards that would drive production and serve as the vision document for the project. These gave us the scenes, what elements would be in each, and how they would interact – snapshots of the action along a timeline.

Like in the physical world, in 3-D there are often many ways of building an object; the trick is to know what it's going to do farther down the track. Does it need to animate? Will it turn inside out at some point? Is it part of a larger system or separate? These and other questions determined how we built each object. We used a 3-D-software program called Discreet Max, which grew out of Auto-CAD 10 years ago. In 3-D space currently, models are built using several different types of algorithms: polygonal modeling, NURBS (nonuniform rational B-spline) modeling, and procedural blobmesh modeling using simple shapes, for example.

Our typical projects range in cost from $10,000 to $150,000. When budgets and time permit, we aim higher, to a level we call faux-realism. The term "photo-real" gives pause to 3-D artists, and is used entirely too often without understanding what is required to get there. It is achievable, but not without a significant amount of time and effort, and most 3-D shops simply don't have the time or manpower to do it. The human eye has the capacity to find the minutest flaw, which immediately blows the "reality" of a scene, so it's a slippery slope.

The animation phase proved the most time-consuming, as is often the case. While choreographing the animation we also applied textures to each model using raster bitmaps and/or procedural resolution-independent textural shaders. Lighting each scene happened concurrently with texturing, since one affects the other, and camera moves were choreographed and tweaked.

Each scene was then rendered out as a series of sequential frames, 30 frames per second, which were stitched together in a compositing program called Adobe After Effects, and then output as final movies suitable for broadcast, DVD and the net. (We supplied them with different resolutions.) The great thing about working digitally is that if and when paradigms change, we can go back and update the 3-D scenes to reflect current knowledge.

Sometimes we find ourselves animating processes where the precise interaction hasn't yet been mapped, as in the case of an animation we did for Nature Reviews Genetics on RNA interference (see box). It is an occasionally frustrating, often intoxicating space – to be at the bleeding edge of what is known. But this is exactly where we want to be.

WHAT IS THE VALUE?

We exhibited at the 2003 AAAS meeting in Seattle where, unbeknownst to us, the last two days of the exhibition were designated as "Family Days." We watched toddlers to octogenarians become totally engaged watching the animations. The children especially were fascinated, holding back until the last moment, grilling us with questions while their parents tried to pull them away. On show-and-tell day at an elementary school in Maryland several years ago, a colleague watched, fascinated, as first graders turned to one another and explained what they were seeing, insisting on watching it seven times. The following year those same children, now in the second grade, had inquired whether they could see that animation again. They had remembered and could explain what they had seen. If those aren't good indications that 3-D animations can spark interest in science, then I'm not sure what would be. It's inspiring.

One objection we've had is that good animations are too "pretty." To that, we say, what's wrong with presenting information in a visually pleasing manner, if it serves to enhance the experience, and remains in the mind of the viewer longer? It's not speculation to say that science has a bad rap – from a life-long experience as a layperson I can safely argue that many people regard science with varying mixtures of boredom, fear, confusion, and envy. That divide must be bridged; science has to become cool again.

Although we have worried in the past about our animations perhaps being too simplistic for the science community, it appears in fact to stimulate conversation between disciplines. They've been used as a tool to encourage greater exchange of ideas. That was an unexpected pleasure, and showed us that these animations can have far greater reach than initially anticipated.

There is, right now, someone out there, very young, very bright, who might see a science animation that turns on a life-long quest to investigate. Watching a mind light up with understanding is one of the greatest joys I will ever know. That's our goal.

Animation should never supplant books. There is a particular joy in reading and assimilating information via text, and that should never become outmoded. Animation can serve to augment that understanding and provide visual anchors. After all, humans have been visualizing for thousands of years. 3-D animation is just a new paintbrush.

However, the last 10 years have brought an exponential increase in the amount and quality of visual information presented to audiences, and they have come to expect it. That's true of all forms of visual communication – not only entertainment programming, but formal and informal education of all stripes. It's the way forward.

Beth Anderson is cofounder of Arkitek Studios http://www.arkitek.com, which develops content for the science, technology, and education communities. Also a musician, she has designed and built biotech instrumentation, and holds two patents with her father on Navstrip™, an aircraft navigational aid, and ClonePick, a film-based antibiotic screening method.

She can be contacted at beth@arkitek.com.